Simulation of Cement Reactivity with CO2 in the Wellbore Environment
J. William Carey, Peter Lichtner, and Chuan Lu Los Alamos National Laboratory
2007 Wellbore Integrity Network Workshop (Santa Fe)
Acknowledgements: DOE’s National Energy and Technology Laboratory (04FE04-07)
LA-UR-06-0636
Motivation: Model the long-term integrity of wellbore cement exposed to CO2
• Determine mode of CO2 interaction with cement
• Develop model of changes in effective cement permeability as function of CO2 reaction
• Need initial conditions of interfaces (width, porosity, effective permeability)
• Need initial drive (diffusion, buoyancy, capillary, gradient)
• Calculate changes in interface permeability
• Ultimately, couple with geomechanics
Casing Shale
Grout-CasingInterface
Hydrated Cement
Grout-ShaleInterface
Matrix Diffusion
Interface Flow Interface Flow
Fracture Flow
• How does the two-phase system affect reaction rates? Will CO2 penetrate cement primarily by diffusion, pressure-driven flow, or capillary pressure? LA-UR-06-0636
Effect of Two-Phase Behavior on CO2 Reactivity with Cement
1-D Calculations• Role of capillary pressure properties of cement,
shale, and reservoir rocks • Comparison with no-flow (diffusion) results
2-D Calculations• Problem set-up and boundary conditions• Preliminary results for flow-only case
LA-UR-06-0636
0 0.05 meters 0.25
FormationCement
38% - - C S H(x 2SiO =0.36, / Ca Si= 1.78)
15% portlandite
14% monosulfate
3% hydrogarnet
30% porosity
20% illite
7% quartz
1% kaolinite
1% calcite
1% dolomite
70% porosity
• 1- D diffusion of CO2- saturated brine intocement
• 25 o 179 (C and bars P CO2)• : , , , Variables Porosity tortuosity reaction rates
and solid solution model [ & (2007) Carey Lichtner American ]Ceramic Society
LA-UR-06-0636
1-D Diffusion of CO2-Saturated Brine
LA-UR-06-0636
Two-Phase Simulation of Cement-CO2 Reaction
• 1-D (geometry and mineralogy as before)• Use parallel version of FLOTRAN (PFLOTRAN)• Parameters:
– As before (reaction rates, porosity, tortuosity)– Without solid solution model– Need permeability and relative permeability (capillary
pressure relations)
• Horizontal geometry (as at SACROC)– CO2 interaction by diffusion and capillary pressure
• Scoping calculations
LA-UR-06-0636
Capillary Pressure Relations
(Bennion and Bachu, 2006, SE 99325Savage and Janssen, 1997, ACI Mat. J.)
LA-UR-06-0636
Initial Conditions
• T = 50 C; P = 200 bars• Cement: saturated with 1.6
M NaCl brine• Formation: 50% saturated
with brine/CO2
Perm (mD) alpha m Φ τ
Cement 0.1 0.0023 0.42 0.4 0.005
Shale 0.1 0.01 0.9 0.2 0.5
Reservoir 100 10 0.5 0.2 0.5
Ψ = (S-1/m – 1)(m-1) / α
LA-UR-06-0636
Cement: Reservoir (Flow Only)
LA-UR-06-0636
Cement: Reservoir (Flow Only)
LA-UR-06-0636
Cement: Shale (Flow Only)
LA-UR-06-0636
Cement: Shale (Flow Only)
LA-UR-06-0636
No Contrast: Cement-like Properties
LA-UR-06-0636
No Contrast: Cement-like Properties
LA-UR-06-0636
Cement: Reservoir (1 year)
LA-UR-06-0636
Cement: Reservoir (30 years)
LA-UR-06-0636
Cement: Cement-like (1 year)
LA-UR-06-0636
Cement: Cement-like (30 years)
LA-UR-06-0636
2-D, Two-Phase Wellbore Problem
• Narrow, high permeability zone at cement-caprock interface
• Cement and caprock have similar capillary pressure functions
• Buoyancy driven CO2 movement (no pressure gradient)
• Minimum CO2 plume thickness required to overcome capillary barrier
• Constant pressure along top and bottom boundaries (hydrostatic-like pressure gradient)
LA-UR-06-0636
CO2 saturation at 0.05 years
LA-UR-06-0636
CO2 saturation at 0.2 year
LA-UR-06-0636
Dissolved CO2 at 0.2 year
LA-UR-06-0636
Conclusions
• Capillary pressure-driven drainage of cement can result in more rapid CO2 penetration into cement
• Supercritical CO2 along interfaces or other high porosity (low capillary pressure) contacts has low capillary driving force
• Capillary driven flow of CO2 into high-quality cement unlikely
• High water/cement ratio promotes low capillary pressure properties and may allow capillary sorption of supercritical CO2
• 2-D simulations allow investigation of interface flow dynamics; relative importance of drive by pressure gradient versus buoyant flow unclear
LA-UR-06-0636
Top Related